Mapping Environmental Barriers to Arbovirus Outbreaks in Brazil

How climate variables affect the distribution of mosquito-borne viruses?

🧩 Puzzle

  • The 2016-2022 Yellow Fever (YF) wave was unprecedented. It breached historical barriers, spreading from the Amazon through the Cerrado and into the densely populated Atlantic Forest.

  • Critical questions:

    • How did the virus persist through temperate winters and high-altitude regions (< 16°C) that should have stopped transmission?

🌡️ Hypothesis

  • We propose that anomalous weather conditions (climate change & variability) created temporary bridges for viral survival.
    • Warmer Winters: Prevented viral die-off in historically cold zones.
    • Altered Rainfall: Sustained vector populations during dry seasons.

Background

Temperature anomalies are the biological constraints of the virus. The viral incubation speed depends on temperature conditions (Bellone and Failoux, 2025; Johanssen et al, 2010).

Viral incubation speed.

Background

Relationship between temperature and R0 value for Yellow Fever virus transmission.

🎯 Objectives

  1. Data Integration
  • Building a comprehensive database linking NHP epizootics with fine-scale climate data (Temp, Rain, Wind, Vegetation).
  1. Spatio-Temporal Model
  • Identifying “Thermal Corridors” and analyzing how microclimates allowed viral persistence during winter.
  1. Predictive Forecasting
  • Developing a risk model using climate indicators as “early warning” signals for future outbreaks.

Temperature corridors

Expected minimum temperature for Brazil along the year.

Expected minimal temperature of each municipality in the south east (SE) region of Brazil along the year. The reference temperature was obtained by averaging the weekly minimal temperature between 1986 and 2015.

Climate anomalies during the YF outbreak

Municipalities affected by yellow fever in the Southeast region of Brazil between January 2016 and June 2025 (red line). Municipalities affected by minimum temperature anomalies in the Southeast region of Brazil between January 2016 and June 2025 (dashed blue line). Data were aggregated by epidemiological week. We considered a climatic anomaly as a minimum temperature above 5 standard deviations from the expected average for the 30-year period (1986–2015).

Source: OpenDataSUS; Coopernicus CDS.

Association of affected municipalities and climate anomalies

Number of non-human primate (NHP) yellow fever cases per municipality in the Southeast region of Brazil between January 2016 and June 2025.

Number of temperature anomalies per municipality in the Southeast region of Brazil between January 2016 and June 2025.

Scientific Challenges

  • ⚠️ Sampling Bias
    • Dead monkeys are only found where people look.
    • Solution: Occupancy Modeling & “Observer Bias” Covariates.
  • ⚠️ Microclimate vs. Macroclimate
    • Satellites miss the warm tree hollows where mosquitoes hide.
    • Solution: Topographic Downscaling & Canopy Adjustment.

Relevance and Impact

  • 🚀 Proactive Surveillance: Moving from reactive monitoring to early warning.
  • 🛡️ Public Health: Enabling vaccination campaigns months in advance.
  • 🌍 Scalability: A framework replicable for WNV and Leishmaniasis.


Thank You!


Questions & Discussion


Mauro MORAIS
Institut Pasteur de São Paulo (IPSP)
mauro_morais@pasteur-sp.org.br